Module 1, Part 1: Introduction to Cloud Computing Services

Course Information and Objectives

• Course Identification:   - Year: $2025-26$   - Subject: Cloud Computing Services ($CCS$)   - Class: BE Semester $VIII$ ($C$)   - Module: $1$, Part $1$   - Instructor: Dr Manoj K Sabnis

• Course Objectives:   - 1.1 Fundamentals: Understand the fundamentals of Cloud Computing including its components, deployment models, service models, and characteristics.   - 1.2 Service Models: Explore and analyze cloud service models such as $SaaS$, $PaaS$, and $IaaS$, along with emerging services under the $XaaS$ (Anything as a Service) umbrella.   - 1.3 Practical Implementation: Gain practical knowledge of $OpenStack$ and Serverless Computing, including architectural components and serverless functions using $AWS\,Lambda$.   - 1.4 Ecosystems and Integration: Comprehend the structure and role of cloud ecosystems, $APIs$, and third-party services integration within cloud environments.   - 1.5 Security: Understand cloud security challenges and solutions, including Identity and Access Management ($IAM$), encryption, audit, and risk mitigation techniques relevant to cloud platforms.

Evaluation Scheme

• Examination Breakdown:   - End Semester Exam: $60$ marks.   - Midterm Test: $20$ marks.   - Continuous Assessment: $20$ marks.

• Continuous Assessment Details:   - Module 1 + Module 2: $10$ marks (Case study: design of one secondary service).   - Module 3 + Module 4: $5$ marks (Assignment).   - Module 5: $5$ marks (Case study: design of cloud security solutions).   - Total Assessment: $20$ marks.

Evolution of Service Modules: From Single Services to Applications

• Single Service Module Architecture:   - Structure: Initially, a single program consists of input, code, and output as a single module.   - Processing: Small modules run on small machines but are centralized for small-level processing.   - Security (Single Service): Object-level programming separates code and data. The user provides data input and receives processed output, while the business logic (code) remains hidden.   - Scaling: Large modules require single servers with higher processing power for single medium processing types.

• Capacity Planning for Single Modules:   - Analysis: Determined by resource requirements (Software and Hardware).   - Parameters: Size, cost, availability, robustness, performance, and accuracy.   - Reliability Metrics:     - $MTTR$: Mean Time To Repair.     - $MTBF$: Mean Time Between Failure.   - Aim: Selection of hardware and software based on a cost-effective and efficient solution.

Client-Server Models and Security

• Basic Security Model: Client-server models rely on object-oriented programming to process code on the server side while maintaining request-response interactions on the client side.
• Client Types:   - Thick Client: Possesses memory and limited processing; allows input to be preprocessed before being sent to the server.   - Thin Client: Only sends requests and receives responses; possesses no memory for storage.   - Mobile Client: Moves with the user, small size, low processing power, more presentable interface, and uses a special Operating System ($OS$).
• Definition of Service Module: A set of inputs processed by code to produce a required output or set of outputs.

Current Scenarios: Service Domains to Enterprise Applications

• Service Progression:   - Service: A set of inputs and outputs.   - Service Domains: Groups of like services (e.g., Financial, Business, Entertainment).   - Application: A collection of different service types/domains.   - Enterprise Level Application: Multiple service domains integrated at a high level.
• Implementation Considerations:   - Deployment: Can be central or distributed, impacting hardware and software layout.   - Scalability: Enterprise applications must be horizontal and vertical scalable in terms of resources.   - Business Aspect (Capacity Planning): Profit involves ensuring Resource Provisioning is less than ($<$) Resource Requirement through management strategies.

Stakeholder Management and System Design

• Management Planning Phases: Requirements, Design, Implementation, Testing, Evaluation, and Feedback.
• Resource Budgeting (Demand vs. Provisioning):   - Status: Resources Required ($>$) Resources Provisioned.   - Critical Resources: Storage (Space), Processing (Compute), and Communication (Bandwidth).
• Data Challenges:   - Data Explosion: Massive volumes of varied data types.   - Volatility: Data useful today may be useless tomorrow (runtime applications).   - Temporal Data: Analysis of past, present, and predicted (tomorrow) data.
• Information Life Cycle ($ILC$): Identify, Acquire, Migrate, Archive, and Destroy.
• Compliance: Establishment of rules and security for hardware, software, and human resources (rules).

Human Resources and Strategic Roles

• Top Management: Decision-making team.
• Managers: Database ($DB$) managers, store managers, and network managers who control activities.
• Implementation Team: Designers, engineers, and network administrators.
• Vendors: Provide hardware and software.
• Service Providers: Offer tailor-made services based on user requirements.
• Brokers: Act as contact persons between parties.
• Consumers: End-users of the services.

Wide Enterprise Level Business Application Requirements

• Compute Power ($Processing$):   - Front-end: User-facing dynamic $GUI$ for queries and visualized data analysis; includes hardware like interactive screens and user-interface software.   - Main Processing: Central or distributed high-speed parallel processing servers.   - Back-end: Database management and data archiving.   - Supporting Features: Load balancing, security, and administration.
• Space for Storage:   - Physical Layers: $RAID$ (Front-end), Hard disks (Middle), and Tape drives (Back-end).   - Software: Storage management and security software.   - Management: Backup facilities, data recovery, protection, and user authentication.
• Bandwidth for Communication:   - Software: Various communication protocols.   - Hardware: Networking/internetworking devices for $LAN$, $WAN$, and $MAN$ infrastructures.

Transitioning to Cloud Service Models

• The Ownership Problem: Owning infrastructure requires high initial investment, significant time/energy for procurement, security maintenance, and server administration.
• The Cloud Solution:   - Managed by a Service Provider instead of the developer.   - Developer focuses solely on application development.   - Resources are provided as needed (Power, storage, bandwidth).   - Billing: Pay-per-use basis.

Infrastructure as a Service ($IaaS$)

• Definition: Infrastructure ready-made and customized on rent on a pay-per-use basis. It is the first basic service provided by the cloud.
• Standard Cloud Components:   - Interconnection of Nodes: Networks divided into manageable $LAN$, $MAN$, and $WAN$.   - Internetworking Devices: Routers and Gateways.   - Networking Devices: Hubs, repeaters, and switches.   - Nodes: Processing, storage, and end-user nodes.
• $IaaS$ Provisioning Process:   1. User lists requirements (processing power, storage space, bandwidth).   2. Requirements are sent to the provider.   3. Provider performs a requirement survey and drafts $2$ to $3$ design plans.   4. User evaluates and selects one plan.   5. A Service Level Agreement ($SLA$) is drafted, detailing resources, budget, payment methods, and timelines.
• Configurations: Standalone, Centralized, Distributed, Grid Computing, and Utility Computing.   - Utility Computing: A business model where users use hardware/software and receive "utility tokens" for usage, benefiting future use and locking in existing users.

Platform as a Service ($PaaS$)

• Requirement: Software development platform, testing, debugging, evaluation, deployment, and interfacing software.
• Problem: Licensing for specialized development platforms is expensive, and tools often are only needed once.
• Solution: Developer provides requirements and timelines; the provider offers the platform and technical support.
• Definition: Softwares and licenses are the provider's responsibility; the developer rents them and pays per use.

Software as a Service ($SaaS$)

• Requirement: Deployment, reaching users, feedback implementation, scalability, and security.
• Solution: Cloud provider handles application deployment, security, advertising, updates, billing, and accounting.
• Goal: Allows the researcher/developer to focus on development rather than administrative and technical support tasks.

Cloud Service Categories

• Primary Services:   - $IaaS$: Hardware infrastructure.   - $PaaS$: Developing platforms and testing software.   - $SaaS$: Application delivery and feedback management.
• Secondary Services: Developer-made applications hosted on the cloud that become services themselves.   - Examples: Database as a Service, Storage as a Service, Security as a Service, Integration as a Service, Collaboration as a Service, Backup as a Service, Disaster Recovery as a Service.

Core Features of Cloud Computing

• Standardization: Use of standard hardware and software to ensure "plug and play" compatibility and fewer interfacing issues.
• Automation: Minimal human intervention to control widespread points, reducing human errors.
• Abstraction: Hiding implementation details. Users see the power/storage but not the layout of the physical machines or network.
• Virtualization: Provides stakeholders with the specific view they need.   - Types: Virtual Disk ($RAID$ views), Virtual $OS$, Virtual Network ($VPN$), and Virtual Memory.

Stakeholders and Functions Table

Advantages of Cloud Computing

• 12.1 On-demand Services: Services are available based on user request regardless of current provider stock.
• 12.2 Broad Network Access: Multiple networks allow for scaling of communication channels.
• 12.3 Resource Pooling: Dynamic allocation and deallocation; resources are never idle and are reassigned based on timeline need.
• 12.4 Rapid Elasticity: Support for upscaling and downscaling (Horizontal and Vertical).
• 12.5 Measured Service: Usage is measured, audited, and reported.
• 12.6 Lower Cost: Cost reduction strategies for processing and storage.
• 12.7 Vendor Independence: Standardization and modular architecture reduce lock-in.
• 12.8 Ease of Utilization: Supports Bring Your Own Device ($BYOD$).
• 12.9 Latest Updates: Document sharing ensures only the final updated copy is accessible from any location.
• 12.10 Single Sign-On ($SSO$): Users log in once and move across networks without re-authentication.
• 12.11 Pay-per-use: Payment is strictly based on the duration of usage (e.g., booked for $7$ days but used for $3$, pay for $3$).
• 12.12 No Human Intervention: Avoids human errors through automated control.
• 12.13 Trust and Faith: Established through the $SLA$.
• 12.14 Comprehensive Security: Coverage for data, information, network, and infrastructure.
• 12.15 Layered/Modular Approach: Modules can be added or removed as needed.
• 12.16 Multitenancy: Efficient resource relocation (e.g., User 1's freed resource is given to User 2, potentially lowering User 1's bill).

Disadvantages of Cloud Computing

• Downtime: Significant risk as businesses depend entirely on internet access.
• Vendor Agreement: Potential difficulty in shifting platforms once dependent on a specific vendor.
• Limited Control: Users have little freedom; resources are provider-controlled.
• Security Risks: Sensitive data is managed by third parties; unresolved or unused nodes are vulnerable.
• Internet Dependency: The internet is the absolute backbone of the infrastructure.
• Cost for Providers: While low cost for users, providing high-quality $IaaS$, $PaaS$, and $SaaS$ is expensive for the provider.
• Technical Issues: Complexities arising from distributed infrastructure.

Deployment Models

• Private Cloud: Dedicated to a single organization (Primary + Secondary services).
• Public Cloud: Services available to the public; accessed over the internet.
• Community Cloud: Shared by multiple organizations with common concerns (Shared storage/servers).
• Hybrid Cloud: Integrates Public and Private clouds. Access to authorized persons for private sections while allowing public access to others.

The Cloud Cube Model

• Definition: A four-dimensional model used for designing cloud deployment strategies.
• Dimensions:   - 1. Physical Location: Data kept internal or external to the organization.   - 2. Ownership: Proprietary vs. Open (Technology, software, vendors, interoperability).   - 3. Security Boundary: Perimeterized (inside) vs. De-perimeterized (outside).   - 4. Resource Location: Onsite (within the company) or Outsource (external).